There are approximately 100,000 patients with sickle cell disease (SCD) in the United States and 240,000 children born with SCD each year in Africa. Despite regular transcranial Doppler screening, 38% have silent, white matter strokes detectable by MRI before their 14th birthday leading to significant cognitive dysfunction. Similar brain lesions are seen in patients with thalassemia intermedia, obstructive sleep apnea, severe acute anemia, and the elderly, suggesting a general, non-sickle-related, mechanism. We postulate that anemia itself, by diminishing cerebrovascular reserve (CVR), predisposes the brain to injury in key vulnerable areas. Our arguments are based on our following preliminary data: 1) Anemic patients increase their resting cerebral blood flow (CBF) such that oxygen delivery is normal, independent of the anemia mechanism. 2) CVR declines inversely with resting CBF, approaching 0% in some SCD patients 3) we observe white matter loss and abnormal diffusivity in both SCD patients and non sickle cell anemia patients. 4) The pattern of volume loss and abnormal diffusivity predicts regions of low CVR from a completely different patient cohort. We propose to study the relationship between CVR, anemia, and white matter loss in chronic anemia patients at particularly high risk for silent stroke (sickle cell anemia, thalassemia, and other rare anemias). Specifically, we will determine if focal areas of decreased CVR occur in the same distribution as silent strokes and preclinical white matter injury, and whether transfusion improves CVR these vulnerable regions. Increase in regional flow measured by arterial spin labeling MRI in response to acetazolamide is a measure of CVR. Thus the goal of this project is validate CVR as a biomarker for white matter vulnerability that can be used as a means of discovery for future clinical trials of silent stroke prophylaxis. To validate CVR, we will first identify its relationship with hemoglobin level and known vascular biomarkers to identify potentially modifiable risk factors. We will determine whether low CVR predicts decreased white matter volume, loss of axonal integrity, silent stroke, and impaired neurocognitive function in chronically anemic subjects as compared to controls. Lastly, we will test whether raising hemoglobin (simple transfusion), lowering hemoglobin S (isocrit exchange transfusion) or hydroxyurea improve CVR in the white matter regions at risk for stroke. Although targeted toward high-risk groups (hemoglobinopathies), this research has profound implications regarding white matter disease found in the general population, including the elderly, patients with sleep apnea, diabetes, and hypertension. White matter strokes occur in more than 30% of individuals older than 65 and are associated with neurocognitive dysfunction, depression, falling, and all-cause mortality. Anemia is common in the elderly (9.6% of patients in assisted living) and increases white matter stroke risk by 1.8 fold. Thus the techniques and knowledge gained from this study will translate broadly.